Application Research Of 1D-Var Retrieval Technology In Cloud- And Precipitation-effected Satellite Microwave Observation Data Assimilation

The observations used in data assimilation mostly derive from satellite observation which makes up blank regions in oceans and plateaus, and the cloud- and precipitation-affected satellite observations are significant parts of them. However, because of inaccuracy of humidity parameterization process and particle scattering, these observations would be always removed in quality control of data assimilation. Nowadays dominating variational assimilation systems, for example, ensemble assimilation and four-dimension variational assimilation, mainly focus on clear-sky satellite. Development of cloud- and precipitation-affected satellite microwave observations is at start-up phase.ECMWF put forward 1D+4D-Var and all-sky systems one after another in recent years. The former, firstly, retrieves TCWVs in 1D-Var system which then are put into 4D=-Var system as penalty observations. But its handling method of clear-sky observations is disagreeable with one of cloud-sky and their observation operators are also different. All-sky system is developed for 1D+4D-Var’s drawbacks and directly assimilates clear-sky and cloud-sky satellite microwave brightness temperatures in 4D-Var system which develops analysis of humidity and cloud, and improve accuracy of cloud- and precipitation-affected regions’ numerical prediction.My paper does some search about 1D-Var retrieval of 1D+4D-Var system, that builds cloud-effected microwave satellite data retrieval platform by adding super-saturation penalty part in cost function and super-saturation checking in background profiles. After configuring related radiance transmission coefficient files and parameters about SSM/I instrument, add nonlinear restraint factor in background error covariance matrix’s specific humidity part. Contrast experiments prove that nonlinear restraint factor could reduces nonlinear effort from observation part of cost function effectively and adding super-saturation course could make retrieved specific humidity analysis profile correspond better with real physical profile.In water vapor retrieval experiment, adding force-cloud restraint course makes retrieved liquid water path more than zero all the time. Contrasting to clear-sky 1D-Var retrieval experiment, retrieved liquid water path system bias from force-cloud restraint system could be neglected and retrieval effect is better. Afterwards computing retrieval deviation distribution through real LWP and retrieved LWP verifies sensibility relation between LWP and brightness temperature.